Details of Award
NERC Reference : NE/Y003632/1
Sampling The Environment: Model And Design-based Sampling And Data Analysis
Training Grant Award
- Lead Supervisor:
- Professor RM Lark, University of Nottingham, Sch of Biosciences
- Grant held at:
- University of Nottingham, Sch of Biosciences
- Science Area:
- Atmospheric
- Earth
- Freshwater
- Marine
- Terrestrial
- Overall Classification:
- Atmospheric
- ENRIs:
- Biodiversity
- Environmental Risks and Hazards
- Global Change
- Natural Resource Management
- Pollution and Waste
- Science Topics:
- Statistics & Appl. Probability
- Abstract:
- There are many challenges in environmental science and management for which it is necessary to collect data which are then used to make inferences about processes in the environment (e.g. is vegetation cover changing over time?) or decisions about interventions (e.g. is it necessary to undertake soil remediation before a change of land use?). Even in an era where earth observation and sensor networks provide much information, the choice of where to make direct observations (e.g. collect samples of soil, water, air pollutants), when to make them (to monitor change) and how to analyse the resulting data are critical. This process is sampling, a branch of applied statistics. The choice of sampling design can make the difference between largely wasted field effort, and unsafe conclusions, and the efficient use of time and resources to support inferences and decisions about complex processes. Sampling design is undertaken at different scales. It is applied in the development of national-scale surveys or monitoring networks, or even global surveys. These may entail the setting up of sensor arrays (such as the COSMOS sensors used to measure soil water content in the UK), or the running of household surveys to understand environmental impacts on health, livelihoods and wellbeing. Sampling may be undertaken over smaller areas, e.g. at catchment scale, or over managed fields. In all cases a sound design is required, and one which is matched to the problem and the objectives of the survey. A sound understanding of sampling enables the scientist to make good decisions in new sampling tasks, and equips them to discuss challenging problems with statisticians and to recognize when such input is required. However, it is also important to understand sampling issues when undertaking secondary analysis of existing data sets. These include the many large and important data sets held by NERC on things as diverse as the baseline geochemistry of the UK, ground water, space weather and biodiversity. All of these sets where collected according to some design, and this limits how the data can be validly analysed for future use. This course will address the sampling problem with a balance between theory and principles and the practicalities of sample design and data analysis. It will draw on the experience of the University of Nottingham and British Geological Survey in undertaking sampling of water, soil and sediments over a range of scales, and in combination with sensors and earth observation. It will address the distinction between classical "design-based" survey to estimate population parameters through random independent selection of sample sites, and "model-based" methods better suited to tasks such as spatial mapping, and the analysis of data collected at regular intervals from sites or samples. The course will be undertaken on the open-source R platform, which means the tools it will be present will be free and accessible to all. The presenters have strong academic backgrounds in statistics and environmental science, but also practical field experience and experience of survey design at national scale in the UK and overseas for varied tasks, both to support research, policy and management. This will be reflected in the material and examples provided.
- NERC Reference:
- NE/Y003632/1
- Grant Stage:
- Completed
- Scheme:
- Doctoral Training
- Grant Status:
- Closed
- Programme:
- Advanced Training
This training grant award has a total value of £43,845
FDAB - Financial Details (Award breakdown by headings)
| Total - Other Costs |
|---|
| £43,845 |
If you need further help, please read the user guide.